Shell Global Solutions

TRANSFORMING REACTOR
PERFORMANCE
How the BP-Husky Toledo refinery unlocked high
performance from its ageing hydrocracker with a series of
reactor internal design innovations

A WHITE PAPER CO-DEVELOPED WITH BP-HUSKY TOLEDO

Shell Global Solutions

CONTENTS
1. Project objectives 3

2. Project drivers 4

3. Vendor selection 6

4. The challenge of revamping reactors 7

5.
Technology highlights 8

6. How the revamp improved safety 12

7. How the revamp reduced shutdown time 13

8. Results 14

9. Installation and start-up 16

10. Key takeaways 17

2
 Shell Global Solutions

1. PROJECT OBJECTIVES
The 160,000-bbl/d BP-Husky Toledo refinery in the USA, which BP operates as part of
a joint venture with Husky Energy, is a highly flexible, highly complex facility.

However, in the years leading up to 2013, several performance constraints had affected the refinery’s
principal conversion unit, the 31,000-bbl/d hydrocracker. This unit processes 18 different feeds from
the refinery, ranging from naphtha to vacuum intermediate gas oil (its heaviest feed) and cracked stocks.
Having been built in 1964, the hydrocracker was designed to process much-easier feeds than this with
a much less active catalyst, so was ill-equipped to handle the current feeds that the refinery needed it to
process.

The feeds that it was processing had become more difficult to treat and the process conditions more
severe, so that pressure differentials (ΔP) and radial temperature differentials (ΔT) were curtailing the
catalyst cycle length and causing potential safety concerns.

So, in 2013, the refinery’s technologists worked with Shell Global Solutions on a project to address
these constraints. Together, they removed the internals from two reactors (all the existing hardware),
reconfigured them and installed latest-generation internals.

Replacing the reactors was not considered viable. Two other BP refineries, Texas City and Cherry Point,
had achieved impressive results by revamping reactors through installing Shell reactor internals, so
refinery management did not believe that high-capital expenditure replacement would be necessary.

As this paper explains, this was a very technically demanding project that required the challenges
inherent in revamping reactors to be overcome.

3
Shell Global Solutions

2. PROJECT DRIVERS
As shown in Figure 1, the hydrocracker has two five-bed reactors: a first-stage (hydrodenitrification)
reactor, followed by a second-stage (cracking) reactor. Both reactors were revamped, but they were
tackled as separate projects with their own individual commercial drivers.

The first-stage reactor was limited by:

pressure differentials (ΔP) across the reactor, which had been affecting
performance, as the technologists were having to cut the feed rate or the
hydrogen when the limit was reached;

radial temperature differentials (ΔT) across the beds, which were

limiting catalyst cycle length; and

loss of catalyst utilisation, as indicated by the challenge of utilising the

bottom of the reactor, resulting from high radial ΔTs and channelling.

The second-stage reactor was able to achieve a 10-year run length but its key
constraint was radial ΔT spreads, which were causing:

safety concerns. During the previous cycle, the reactor had poor thermal
distribution, which increased the possibility of temperature runaways.

production constraints. The hot spots the reactor was experiencing were
causing excess cracking resulting in light ends production. This was reducing
the product flexibility. The operators had to keep a constant second-stage
feed rate, which prevented them from swinging between the gasoline mode
and the diesel mode.

Therefore, the refinery’s principal stated revamp objectives were to:

extend the first-stage reactor’s cycle length from two/three years to

four years;

mitigate the safety risks in the second-stage reactor by improving thermal

distribution; and

achieve a gas to liquid shift, thereby converting the excess light ends
production to products such as gasoline, kerosene or diesel.

4
 Shell Global Solutions

Fuel gas
C3
Hydrogen make up Hydrogen make up

Feed
Absorber stripper Depropanizer

HDN ISO

IC4/NC4

HPS HPS

Light gasoline
Fuel gas
Heavy gasoline

Kerosene
LPS
Diesel

Stabilizer Recycle splitter

Bleed to FCC feed

FIGURE 1: BP-Husky Toledo refinery’s 31,000 bbl/d two-stage hydrocracker.

5
Shell Global Solutions

3. VENDOR SELECTION
Following a rigorous supplier and technology evaluation, BP selected Shell Global Solutions based on
key criteria including:

■■ track record. BP uses Shell reactor internals in six other hydrocrackers in its refinery
network.
■■ retrofitting experience. The unit’s age would present engineering challenges. Shell
could provide references for other similar projects it had successfully carried out.
■■ reputation. Shell’s work on other projects had been detailed in the design and
engineering, and delivery of parts, and the manufacturing process.
■■ mock-ups. Shell’s procedure of building a full mock-up, a replica of the reactor on
which the contractor can practice assembling the reactor internals, was also a key factor,
as this can facilitate a reactor’s safe and efficient installation. During these mock-ups, the
completeness and correct fit of all parts can be established, which helps to ensure that the
turnaround will not be extended for modifications to be made on-site.

6
 Shell Global Solutions

4. THE CHALLENGE OF REVAMPING VERY OLD REACTORS

The age of the reactors presented a constraint to the project team. When the reactors were built,
the catalysts available were far less active than those currently on the market and the feeds typically
processed were much easier to treat, as they contained less sulphur, nitrogen and foulants. Consequently,
the reactors’ quench lines and thermometry, for example, would not provide all the information currently
required for the most-effective processing.

Essentially, therefore, the inside of both reactors had to be updated and the engineers had to devise
bespoke mechanical innovations to a series of individual challenges.

Chief among these was that the reactors only had nozzles at the top and bottom for the insertion of
thermocouples and through which cooling gas would enter. Modern reactors are fitted with a series of
nozzles down the side, but new nozzles could not be added by welding. The site had a strong desire
to avoid welding anything to the reactor walls, as is the case with most existing reactors, in case the
reactors’ interior lining cracked. The Shell design enables installation of quench lines without welding,
but it still meant that Shell’s engineers had to create whole systems with passages and thermal expansion
loops to be able to cope with the long vertical lines that were required inside these 27.5-m-tall vessels.

The quench lines had to pass through the catalyst beds. Without line insulation, the quench gas heats
up when passing through the hot catalyst beds, but insulating quench lines introduces other difficulties.
This was a common situation for the project team. For every solution, there were possible complications,
which the team resolved by drawing on experience from revamping similar units.

Other actions taken here include:

■■ adapting the thermal expansion loops by making them more sophisticated and moving
them from the reactor beds to the reactor dome;
■■ replacing the thermobars with flexible thermocouples in a smart layout that offered radial
coverage;
■■ removing the internal catalyst bed dump tubes but retaining the penetrations in the bed to
help avoid the gas and liquid bypassing the interbed reactor internals, which can cause
radial temperature maldistribution; and
■■ moving the quench pipes from the bottom of the beds into the interbeds. In-bed quench
pipes can introduce ΔTs, so moving them further down provided more even mixing and,
therefore, enhanced temperature control.

7
Shell Global Solutions

5. TECHNOLOGY HIGHLIGHTS
The revamp involved removing all the reactors’ existing hardware and replacing it with Shell Global
Solutions’ latest-generation reactor internals, some of which are described below.

5.1. Shell filter trays

As the feed contains higher levels of particulates than the existing catalyst system could handle, it was
important to prevent foulants from reaching the catalyst beds. Achieving this would help to address the
pressure drop issues that were constraining its performance. Shell filter trays (Figure 2) were therefore
installed. These are designed to:
■■ limit fouling by trapping foulants before they enter the catalyst bed, which leads to a
slower increase in pressure drop and a consequent longer catalyst cycle length;
■■ break up the liquid stream from the inlet device; and
■■ pre-distribute liquid and gas to the Shell high-dispersion (HD) tray below.

Shell filter trays offer additional filtering capacity by trapping contaminants and sediments, and have the
benefits of enabling a higher catalyst loading capacity by minimising the height of the top-bed grading
layers and significantly delaying any reactor pressure drop increase due to fouling.

FIGURE 2: The Shell filter tray. The filter elements of the filter tray are filled with the smallest
(spent) catalyst particles in the reactor to enable filtration of the tiniest particles. A recent
development in easy-to-open filter elements enables a significant reduction in tray cleaning
and maintenance time.

8
 Shell Global Solutions

5.2. Shell HD trays

Radial ΔTs had been a key constraint in the first-stage reactor and were limiting the catalyst’s cycle length.
The unit’s original trays were not providing high uniformity of vapour–liquid distribution; this can lead to
undesirable radial temperature maldistribution.

Shell HD trays (figures 3 and 4) feature customised nozzles that use the gas flow momentum to disperse
the liquid as a mist. This differentiates Shell’s technology from conventional downcomers or bubble caps
because the nozzles uniformly wet the entire catalyst bed surface and ensures 100% utilisation of the
catalyst bed (Figure 5).

FIGURE 3: A Shell HD tray. FIGURE 4: A 3D view of a Shell HD tray installed at the

top of the reactor beneath a Shell filter tray.

Conventional tray Bubble-cap tray Shell HD tray

10–15% wetting 20–30% wetting Near-perfect wetting

Inert material Dry catalyst Wet catalyst

FIGURE 5: Shell HD trays help to optimise catalyst utilisation by achieving enhanced vapour–
liquid and thermal distribution. These trays are extremely efficient; they enable nearly 100% of
the catalyst inventory to be utilised and offer high feed rate flexibility.

9
Shell Global Solutions

5.3. Shell ultra-flat quench (UFQ) interbed internals

To achieve ultra-uniform temperature distribution in the reactor, which is imperative for optimum catalyst
utilisation, Shell UFQ interbed internals (Figure 6) were installed. These integrate fully with Shell HD
trays and help to reduce the radial temperature gradient, typically by a factor of 3–5, compared with a
conventional mixer.

Shell UFQ interbed internals provide discrete mixing of the gas and liquid phases with consistent
performance independent of the quench load and feature an ultra-flat design that supports another
important objective: maximising the reactor volume.

Generally, this enables more catalyst to be loaded into the reactor. In this instance, however, the freed
space has enabled better quench mixing in the interbeds and the creation of larger manways for safety
reasons (see Section 6).

FIGURE 6: Shell UFQ interbed internals.

10
 Shell Global Solutions

5.4. Shell catalyst support grids

Shell’s catalyst support grids (Figure 7) were key to the success of the project as they:
■■ are designed for large ΔP over the bed. The typical design mechanical load of
a Shell catalyst support grid is 6 bar (87 psi), which is a robust design enabling longer
cycles if the pressure drop across a bed increases. It is achieved by reconfiguring the
construction and spreading the load more evenly over the existing support rings.
■■ are designed to simplify maintenance requirements and reduce
turnaround time, for example, the panels and beams are fitted loose (no nuts and bolts
are required);
■■ are designed to improve safety, for example, they have large manway openings
and clever levelling gadgets that help to reduce confined space residence time; and
■■ are extremely robust, so they can accept higher loads than conventional designs.

A key feature of Shell catalyst support grids is their wedge-wire construction, which can support larger
and heavier catalyst beds, resist fouling and prevent catalyst fall-through. Wedge wire offers key
advantages compared with wire mesh: most notably, it is not prone to fouling. The V-shape of the wire
means that it is self-cleaning. Moreover, there are no loose layers of wire mesh, no overlay is required
for the wire-mesh pads and no knitting is required. Wedge wire also lasts up to five times longer than
wire mesh.

FIGURE 7: Shell catalyst support grids.

11
Shell Global Solutions

6. HOW THE REVAMP IMPROVED SAFETY

Safety is the number one priority at the BP-Husky Toledo refinery and was a key driver for the revamp.
Before the revamp, the reactors had various shortcomings that management was keen to address,
especially:
■■ small manways, which threatened personnel safety and prevented quick egress in
emergencies; and
■■ hot spots and maldistribution, which can lead to process safety issues.

The new internals have improved the situation in several ways. For instance:
■■ the manways are larger and allow for the quick egress of people;
■■ confined space residence time is shorter, as maintenance is easier;
■■ there are fewer foreign objects and obstructions in the beds that might introduce
maldistribution and cause channelling;
■■ hot spots in the catalyst bed have been significantly reduced and are now well below the
recommended limits; and
■■ hot work is not required. By using wedge pins and split keys (Figure 8), the new internals
can be installed without welding and opened and closed using just a hammer.

FIGURE 8: One of the key design features of Shell Global Solutions’ reactor internals is the use of wedge pins
and split keys, which improve safety and enabler faster installation and shorter turnarounds

12
 Shell Global Solutions

7. HOW THE REVAMP REDUCED SHUTDOWN TIME

Having seen the advantages of the wedge-pin design in other reactors in the BP refinery network that use
Shell reactor internals, technologists at the BP-Husky Toledo refinery believe that this feature will reduce
the catalyst change-out time by 2 d per event.

The new reactor internals have brought other maintenance advantages, too, such as shorter catalyst
loading and unloading time, and simplified maintenance procedures. In addition, the confined space
residence time has been reduced, which will help to mitigate the safety risks during turnarounds.

13
Shell Global Solutions

8. RESULTS
8.1. First-stage reactor
Key improvements from the first-stage reactor revamp are described below:

Reduced ΔP
As shown in Figure 9, the ΔP was about 20–30% lower after the Shell reactor internals
were installed. In addition, the total ΔP limit across the reactor was increased. In the run before
the unit was revamped, the ΔP limit was reached at 650 d.

The first run after the revamp was still not limited by ΔP after 1,100 d. Reducing the ΔP
limitations has enabled the unit to run longer on each catalyst cycle and process more barrels
before a catalyst change-out is necessary.

The ΔP has been slowed down significantly and will no longer constrain the unit’s cycle life to
three years.

Reduced radial ΔT
Despite the increased number of temperature measurement points throughout the reactor, the
radial temperature readings showed a substantial decrease of liquid and gas
maldistribution, as shown in Table 1.

Improved catalyst utilisation

The key benefits of the reduction in radial ΔT are improved catalyst utilisation and reduced
catalyst deactivation. At the BP-Husky Toledo refinery, this has helped to extend catalyst
life by a year through the improved catalyst wetting and increased fouling abatement
performance provided by the filter trays.

Enhanced catalyst cycle length

BP-Husky Toledo refinery’s goal of extending the catalyst cycle length of the first-stage reactor
from three years was achieved. It is now on course to achieve a run of over four years.

Reduced turnaround duration

Turnarounds will be 2 d shorter. This is due to the wedge pins, which remove the need for
cutting and welding.

Increased catalyst life

Moving a quench line in the first-stage reactor unlocked a six-month increase in catalyst
life. The increased catalyst life could enable the BP-Husky Toledo refinery to eliminate a
turnaround every eight years. So, in that period the refinery could benefit from:
■■ 30 d longer on stream;
■■ reduced procurement costs, as one less catalyst batch would be required; and
■■ reduced maintenance costs, as one less turnaround would need to be executed.

If the catalyst achieves a five-year run length, it would do so on one and a half fewer catalyst
batches. Crucially, it would then match up with the second-stage reactor’s 10-year cycle,
during which time only two turnarounds would be required, rather than three required
previously, and there would be 30 d more on stream.

14
 Shell Global Solutions

Reactor ∆P Normalised reactor ∆P

181 181
161 161
141 141
121 121
101 101
∆p, psi

∆p, psi
81 81
61 61
41 41
21 21
1 1
Ju

Fe

Ju

Fe

Ju

Fe

Ju

Ju

Fe

Ju

Fe

Ju

Fe

Ju

O
n

n
ct

ct

ct

ct

ct

ct

ct

ct
b

b
14

15

16

17

14

15

16

17
15

16

17

15

16

17
14

15

16

17

14

15

16

17
Prior to Shell tray installation After Shell tray installation Prior to Shell tray installation After Shell tray installation

FIGURE 9: After installing the Shell reactor internals, the ΔP in the first-stage reactor fell by about 20–30%.

Bed 1 Bed 2 Bed 3 Bed 4 Bed 5

Top of bed Before revamp 2.8 1.7 5.6 2.2 6.1
radial ΔT (°C)
After revamp 0.5 1.4 2.0 4.0 1.7
Bottom of bed Before revamp 4.4 18.3 4.4 4.4 3.9
radial ΔT (°C)
After revamp 2.7 1.6 1.5 3.6 2.5

TABLE 1: The revamp has substantially decreased the radial temperature maldistribution in the first-stage reactor.

8.2. Second-stage reactor

Key improvements from the second-stage reactor revamp are described below:

Reduced radial ΔT
As with the first-stage reactor, maldistribution of gas and liquid has been reduced.

Improved safety
The risk of possible temperature runaways has decreased.

Improved production flexibility

The increased catalyst utilisation and improved ΔTs have enabled the refinery to shift from
vapour to liquid products.

15
Shell Global Solutions

9. INSTALLATION AND START-UP

The whole project took 10 months and the unit started up successfully, on time and within
budget in 2014.

Refinery technologists attribute the success of this to a range of factors, especially the multiple inspections
and sign-offs from Shell’s on-site support (day and night shift) as the internals were being installed. This
procedure involved the refinery signing off each bed for release for catalyst loading before Shell Global
Solutions signed off on the correct installation in accordance with the intended process design.

In such a project, it is crucial that the new internals are installed correctly the first time and that rework is
not required, so that turnaround time is minimised and the project remains on schedule. Thus, although
there was redundancy in this approach, this was deemed acceptable given the high strategic importance
of executing the project smoothly and on time.

16
 Shell Global Solutions

10. KEY TAKEAWAYS

Drivers
■■ The hydrocracker at the BP-Husky Toledo refinery experienced operational
problems associated with its age.
■■ It was built in the 1960s when the feeds were easier to process and catalysts
were less active.
■■ By 2013, ΔP and ΔTs were curtailing its catalyst cycle length and causing
potential safety concerns.

About the project

Management initiated a unit revamp, which involved:
■■ removing the reactor internals;
■■ reconfiguring the reactors; and
■■ installing Shell Global Solutions’ latest-generation internals.

Key results: First-stage reactor

20–30% reduction in ΔP, which increased the cycle length by one

and a half years

Reduced radial ΔT, which improved unit safety

Improved catalyst utilisation, which extended catalyst life by a

year

Relocated a quench line, which unlocked a six-month increase in

catalyst life

Key results: Second-stage reactor

Reduced radial ΔT

Decreased risk of temperature runaways

Improved flexibility to shift from vapour to liquid products

Value delivered
The unit’s age meant that this project was highly technically demanding from the
engineering perspective. Nevertheless, these technical challenges were overcome for key,
high-value results.

17
Shell Global Solutions

ABOUT THE AUTHORS

Sarah McFarland is a Commercial Optimisation Engineer performing retro-analysis at the BP-Husky
Toledo refinery. She has been with BP for 13 years and has eight years of hydroprocessing experience
ranging from process engineering to project development, execution and commercial optimisation.

Mallory Taylor is the Process Engineer for the hydrocracker at the BP-Husky Toledo refinery. She has
been with BP since 2012 and is the technical contact for operations. Mallory has been supporting the
unit operation with performance monitoring and daily problem solving since the last start-up.

Pankaj Desai is the Licensing Sales Manager in the Americas and Global Business Manager for
Hydroprocessing Reactor Internals for Shell Global Solutions (US) Inc. He has 37 years of professional
experience with refining and petrochemical process technologies, catalysts, reactor internals and other
proprietary hardware, and software such as process technology models.

Edwin Maas is the Global Technology Manager of Hydroprocessing Reactor Internals at Shell Global
Solutions (US) Inc. He has 38 years’ experience in the industry and has been a subject matter expert at
Shell for 14 years.

Julien Sigaud is a Senior Hydroprocessing Technologist at Shell Global Solutions International BV. He
is responsible for designing Shell reactor internals, providing support to customers having issues with the
operability of their reactors and supporting Shell’s reactor internals licensing business.

18
 Shell Global Solutions

ABOUT SHELL GLOBAL SOLUTIONS

Shell Global Solutions provides technical consultancy and licensed technologies for the Shell
Group and third-party customers within the energy industry. Shell Global Solutions strives to
deliver innovative technical solutions and effective technology to support its customers in their
day-to-day operations and delivery of strategic plans to improve the capacity and performance
of existing units; integrate new process units into existing refineries and petrochemical
complexes; incorporate advanced proprietary catalyst systems (Criterion) and reactor internals;
through to the design of grassroots refineries.

Shell Global Solutions is affiliated with Shell’s catalyst companies, which innovate and
sell catalysts through a network that includes Criterion Catalysts & Technologies, Zeolyst
International, CRI Catalyst Company and CRI Leuna (formerly known as Kataleuna).

For further information, please visit our website at www.shell.com/globalsolutions.

ABOUT BP-HUSKY TOLEDO REFINERY

■■ Owner: BP-Husky Refining LLC, a joint venture between BP and Husky Energy, operated by BP
Products North America
■■ Location: Oregon, Ohio, USA
■■ Capacity: 160,000 bbl/d
■■ Key units: 35,000-bbl/d delayed cokers, 55,000-bbl/d fluidised catalytic cracking unit,
31,000-bbl/d hydrocracker, 42,000-bbl/d catalytic reformer
■■ Crude slate: Heavy, high total acid number (TAN), sweet and synthetic crudes
■■ Product slate: Gasoline, diesel, jet fuel, propane and asphalt

For further information, please visit our website at www.toledobp.com.

19
Shell Global Solutions is a network of independent technology companies in the Shell Group. In this material, the expressions “the Shell Group”
and “Shell Global Solutions” are sometimes used for convenience where reference is made to these companies in general, or where no useful
purpose is served by identifying a particular company.

The information contained in this material is intended to be general in nature and must not be relied on as specific advice in connection with any
decisions you may make. Shell Global Solutions is not liable for any action you may take as a result of you relying on such material or for any
loss or damage suffered by you as a result of you taking this action. Furthermore, these materials do not in any way constitute an offer to provide
specific services. Some services may not be available in certain countries or political subdivisions thereof.

Copyright © 2018 Shell Global Solutions International BV. All rights reserved. No part of this publication may be reproduced or transmitted in any
form or by any means, electronic or mechanical including by photocopy, recording or information storage and retrieval system, without permission
in writing from Shell Global Solutions International BV.